ASAN Medical adopts GenAI in predicting IDH mutations for Glioma diagnosis
Data was gathered from FLAIR MRI images of 565 glioma patients diagnosed between 2008 and 2020.
A research team from the ASAN Medical Centre in South Korea has adopted generative AI (GenAI) to improve the accuracy of predicting Isocitrate dehydrogenase (IDH) mutations for glioma diagnosis.
The team utilised GenAI-based augmentation (GAA) to train contrast-enhanced T1 and FLAIR MRI images of 565 glioma patients diagnosed between August 2008 and September 2020.
In line with this, data from 119 and 108 patients resulted in an accuracy rate of 93.8% and 83.3% for internal and external validation, respectively.
According to the centre, the GAA generated similar images of glioma with high realism and diversity, whilst the prediction accuracy for IDH mutation was optimised when the number of generated images reached 110,000.
Currently, the study’s findings are published in a journal entitled “Neuro-Oncology.”